Kartik Sheth (Caltech /SSC)

1
Evolution of Barred Spirals in COSMOS Watching
Measuring the Assembly of the Hubble Sequence

• Kartik Sheth (Caltech /SSC)
• Eva Schinnerer (MPIA), Peter Capak, Nick Scoville
  (Caltech), Bahram Mobasher (STScI), Armando Gil
  de Paz (OCIW), Peter Teuben (UM)

2
How did we get here?
Hubble (1929)
Hubble (1936) Amazon.com Sales Rank 1,103,080
3
You wanna go where everybody knows your name

• 27.6 SAB
• 36.8 SBs
• (RC3, De Vaucouleurs '63)

• Bar fraction must increase in the infrared but by
  how much ? (Thronson et al. '89 Block
  Wainscoat '91)
• Menendez-Delmestre et al. Using 2MASS K-band
  58 bars 21 candidates
• Eskridge 00 using H-band images 56 strong
  bars 16 weak bars

4
Using Bars as Signposts of Galaxy Formation
A "troublesome" mode (Kalnajs '72)
Bars always form in cold, massive,
rotationally-supported disks (Ostriker Peebles
'73) Even in disks originally dominated by dark
matter - bars form (Athanassoula 02)

• Epoch of Bar Formation lt-- Presence of cold,
  massive,
• rotationally-supported disks

5

• Bars play a critical role in galaxy evolution
• Drive 108 M? of gas to center (Sakamoto et al
  99 Sheth et al 04)
• Trigger circumnuclear starbursts (Sersic
  Pastoriza 65 --gt Ho et al. 97)
• Feed the central black hole? (Shlosman et al 88)
• Reduce the chemical abundance gradient (e.g.,
  Martin Roy 94)
• May form bulges
• Bar bulge instability (Stanek, Somerville,
  Klypin 03)
• Gas transport -gt psuedo-bulges? (Friedli Benz
  95 Norman et al. 96)
• May ultimately commit suicide! (Norman et al.
  96)
• And regenerate w/ accretion? (Bournard Combes
  02)

Bar destroyed
Bulge forms
6
A Stunning Surprise!
The only example of a possible barred spiral in
the HDF - van den Bergh 96
Further analysis (I-band) showed a paucity of
bars at z gt 0.5
Similar results from CFGRs (van den bergh et al
00)
Abraham et al. 99
7

• Implications (Abraham et al. 99)
• Galactic disks not massive yet
• Massive DM halos
• Disks were dynamically hot
• Bars were being destroyed faster at high-z (by
  mergers!)

Another (less pleasant) possibility - Revisiting
Bar formation?
8
At z1 The 1.6m light is coming from rest-frame
I-band (800 nm) Remember that bars are most
easily identified in the infrared
9
6.3 Gyr ago and earlier

• We detect
• 4 bars 5 candidates
• Previously 2 bars were detected
• Two candidates at z1.66, z2.37
• Only a handful of bars - reflective of a true
  decline?

Sheth et al. (2003)
10
Resolution, resolution, resolution
Sheth et al. (2003)

• At z gt 0.7
• WFPC2 0.8m -marginally capable of detecting a
  typical bar
• NICMOS - even poorer resolution
• Average size of four bars we detect 12 kpc
• Note the possibilities w/ ACS/CARMA eventually
  ALMA

11

• Keeping these caveats in mind, compare bar
  fraction
• 4/95 galaxies at z gt 0.7
• 3/31 if we apply previous I-band cutoff
• 2/9 in 0.8m WFPC2 sample
• At z gt 0.7, (4/95) local fraction (1/44)
• Remember that all selection effects give us a
  lower limit!

No evidence for a declining fraction of barred
spirals at z gt 0.7, as previously claimed
12

• Apples to Apples Sample Selection
• Local samples are well-behaved
• Large samples in multiple-l including IR
  available
• Resolution adequate for resolving primary bars
• Hubble type, inclination, distance well-known
• Underlying disk is also (usually) well-imaged
• Multiple methods for bar identification and
  measuring bar properties possible

COSMOS, GOODS, GEMS

• For high redshift samples
• ALL of these conditions become important caveats
• Selection criteria critical in interpreting
  results

Difficult but not impossible to overcome
Requires ACS Resolution
USE COMMON ALGORITHM (e.g., ELLIPSE-FITTING)
TEST Algorithm using MODELS Local data at
multiple wavelengths
13
Bars as Signposts of Galaxy Evolution in COSMOS

• How does the bar fraction vary with redshift?
• Rate of galaxies arriving onto Hubble sequence.
• How do bar properties (a, e) vary with redshift?
  
• Bar destruction/re-formation
• Interaction of the DM halo the disk
• What is the correlation between bars and bulges
  bars SF ?
• Bars --gt bulges at higher redshifts?
• Gas transport --gt star formation, build up of disk

14

• 2 ACS studies confirm presence of high-z bars
• Bar fraction constant (30) to z1
• Elmegreen 04, 186 galaxies
• Bars were smaller , younger

We investigate the frequency of bars out to z1
drawing on a sample of 1590 galaxies from

• Jogee 04, 258 galaxies
• Matched band observation
• Bars are the same size

15
COSMOS Data
At z lt 0.7, bars are easy to spot
At z gt 1, we are shortwards of the Balmer break
16
COSMOS bar at z1
GALEX bar at z0 (NGC 3627)
17

• Bar identification is very strict (2 signatures)
• Monotonic increase in e drop of d(e) gt 0.1
• AND a constant PA d(PA) gt 10 at bar end.
• Only 1 signature --gt candidate bar

18

• A Preliminary COSMOS sub-sample
• Begin w/ 5-band phot-z catalog
• Choose IAB lt 23 in mapped ACS fields
• Stellarity lt 0.9
• SED type SO -- Irr
• Choose 100 best determined phot-z in Dz 0.2
• Fit galaxies using ellipse-fitting algorithm
• Eliminate high inclination disks, compact
  objects, merging/irreg objects.
• --gt 260 disk galaxies from z0-1.2

• If real, the decline in the bar fraction may be
  indicate the RATE at which disks cooled and
  becoming massive enough to host bars, to
  transform into todays Hubble sequence.

19

• Likely to be a selection effect e.g

20

• Disk size constant w/ time?
• (see upcoming talk by Sargent)
• Little angular momentum exchange w/ DM halo?
• No bar destruction / reformation
• Ellipticity evolution - not possible w/ ACS